2
The Gender Gap in Ideology
Over the past two decades, Americans are increasingly able to identify themselves
as liberals and conservatives, and these self-identifications are increasingly correlated to
partisanship. Many observers see a nation increasingly polarized, engaged in a culture
war (Hunter, 1992; Greenberg, 2004). One of the most striking images after the 2000
election was a map showing counties colored blue if won by Bush, and red if won by
Gore. The map implied a starkly divided nation, with very populous urban strongholds
supporting Democrats and vast stretches of less populated suburbs and rural areas
supporting Republicans.
Yet another statistic tells a very different story of the 2000 election. Gore won
among women by 12 percentage points, and Bush won among men by 10 points.
Although the gender gap in voting was nothing new (Bendyna and Lake, 1993), the 2000
campaign marked the first time that majorities of men and women voted for different
candidates. The gender gap persisted in 2000 despite significant spending by both parties
to shore up their weaknesses with one set of voters or another. Gore wore flannel shirts
and talked about hunting, while Bush emphasized a somewhat undefined compassionate
conservatism.
The emergence and continuity of the gender gap in voting has been the subject of
considerable research (e.g. Sapiro and Conover, 1997). The magnitude of the gender gap
has remained modest when compared with the race gap or the religion gap, and it is
clearly more of a crevasse than a chasm. Yet the gender gap is substantial enough to
affect electoral strategies and election outcomes. Moreover, the gender gap is
theoretically interesting, because men and women often share lives and economic